F13 Raises $5 Million Pre-Seed for AI Vector Graphics

F13, a Berlin-based AI startup, has emerged from stealth with $5 million in pre-seed funding co-led by Credo Ventures and Point Nine Capital, with participation from angel investors including Baobab Ventures' Carles Reina and mainframe's Jack Richardson. F13 is developing foundation models designed to generate precise, controllable vector graphics from text, images, or existing graphics.

Founded by Gregory Janik and Ingimar Tomasson, F13 is targeting a gap in the current wave of AI image generation tools: most popular models produce raster images, pixel grids that look convincing but cannot be cleanly scaled, edited as discrete shapes, or used directly in professional design and engineering workflows that depend on vector formats, such as logos, technical diagrams, icon sets, and data visualizations. Building a model that reliably outputs precise, editable vector graphics rather than pixel approximations is a meaningfully harder technical problem than raster image generation, since it requires the model to reason about discrete geometric structure rather than simply predicting plausible-looking pixel values.

Janik, F13's CEO, previously co-founded SaaS-financing company Vitt and led engineering at education startup Knowunity, while Tomasson also worked at Knowunity and co-authored research presented at the 2025 International Conference on Machine Learning on predicting optimal-transport calculations, a mathematical foundation relevant to precisely mapping and transforming geometric shapes, directly applicable to the kind of controllable vector generation F13 is building toward.

The company's decision to raise without a pitch deck, relying instead on the founders' track record and early technical demonstrations, is a notable departure from typical pre-seed fundraising process, and reflects a pattern increasingly available to repeat or technically credentialed founders in AI, where investors are willing to underwrite a bet on demonstrated capability and a clear technical thesis ahead of a fully built product or formal narrative deck.

If F13 succeeds in reliably generating accurate vector graphics from natural-language prompts, the company would be well positioned to serve design tool makers, marketing teams, and technical documentation workflows currently underserved by raster-only AI image generators, a substantial and currently unmet segment of the broader AI image generation market.

F13's location in Berlin also places it in a city that has quietly built a meaningful concentration of AI startups over the past several years, drawing on a deep pool of technical talent from the city's universities and research institutes alongside an increasingly active local investor base, even as more of the broader AI funding narrative has centered on San Francisco. A pre-seed round of this size, co-led by two established European investors, is a reminder that some of the most technically ambitious AI bets are still being made well outside Silicon Valley.

Founders building in AI-generated design tools or creative infrastructure can find active investors through the match tool, browse AI/ML investors in the directory, or check the live listing of current opportunities. Catch up on more coverage in the newsroom, or register on AngelLinx to start a raise.

AngelLinx Intelligence | angellinx.ai